Metrics
View and query data
The following table describes different metrics VGS provides
Core Platform Metrics
vault_record_usage
(gauge)
tenant_id - vault identifier
env - sandbox / live
action - CREATED / RETRIEVED / DE_DUPE / RE_ALIAS / DE_ALIAS / DELETED
format - Generic VGS alias / Payment Card Format preserved, etc
persistence - persistent / volatile
phase - REQUEST / RESPONSE
route_id - route identifier
Alias interactions performed by proxy: - CREATED - creating an alias to remove sensitive information from the payload and replace them with a synthetic value - RETRIEVED - revealing the alias to restore sensitive data pieces on a previously redacted payload - RE_ALIAS - adding an alias to a multi-alias entry as in different formats of representing the same value (like cc number) - DE_ALIAS - removing an alias from a multi-alias entry - DE_DUPE - attempt to create and redact a record that already exists - DELETED - permanently deleting an alias Rate - the number of record usage, per minute.
vault_record_usage_failure
(gauge)
tenant_id - vault identifier
env - sandbox / live
action - CREATED / RETRIEVED / DE_DUPE / RE_ALIAS / DE_ALIAS / DELETED
error - NOT_FOUND / ACCESS_DENIED / TOKENIZATION_FAILED
persistence - persistent / volatile
route_id - route identifier
Failed alias interactions performed by proxy. Error types: - NOT_FOUND - exception raised if searching for an object and it’s not found - ACCESS_DENIED - exception raised if searching for an object and it’s found but access is denied based on request classifiers and tags on a token - TOKENIZATION_FAILED - exception raised if there are issues during tokenization or detokenization of a value Rate - the number of record usage failures, per minute.
HTTP Proxy Metrics
http_proxy_request
(gauge)
tenant_id - vault identifier
application - http
proxy - reverse / forward
env - sandbox / live
method - http method
Representation of every http request to measure. Rate - the number of requests, per minute, for the services that you are serving.
http_proxy_response
(gauge)
tenant_id - vault identifier
application - http
proxy - reverse / forward
env - sandbox / live
method - http method
code – http response code
Representation of every http response to measure. Rate - the number of requests, per minute, for the services that you are serving.
http_upstream_response
(gauge)
tenant_id - vault identifier
application - http
proxy - reverse / forward
env - sandbox / live
code - http status code
upstream_host - host, proxy upstreamed request to
A counter of the number of upstream responses the proxy has received. Good for rate measurement as much as for errors - the number of failed requests per minute. Rate - the number of requests, per minute, for the services that you are serving.
http_proxy_processing_
duration_ms
(histogram)
tenant_id - vault identifier
application - http
proxy - reverse / forward
env - sandbox / live
method - http method
label - clientToProxyRequest / proxyToClientResponse
Processing time for different phases of requests interacting with proxy. Duration - distributions of the amount of time each request takes.
http_upstream_processing_
duration_ms
(histogram)
tenant_id - vault identifier
application - http
proxy - reverse / forward
env - sandbox / live
upstream_host - host, proxy upstreamed request to
Representation of the time it took to process a request for an upstream. Duration - distributions of the amount of time each request takes.
SFTP Proxy Metrics
sftp_proxy_processing_
duration_ms
(histogram)
tenant_id - vault identifier
application - sftp
env - sandbox / live
sftp_operation - SFTP operation being performed (PUT or GET)
upstream - SFTP upstream host and port
status - SFTP processing status
Processing time for different phases of requests interacting with proxy, with breakdown on SFTP message specifics. Should be used for customers who employ SFTP proxy integration. Duration - distributions of the amount of time each request takes.
sftp_proxy_request
(gauge)
tenant_id - vault identifier
application - sftp
env - sandbox / live
sftp_operation - SFTP operation being performed (PUT or GET)
upstream - SFTP upstream host and port
status - SFTP processing status
Representation of every SFTP request to measure. Rate - the number of requests, per minute, for the services that you are serving.
ISO8583 Proxy Metrics
iso_8583_proxy_processing_
duration_ms
(histogram)
tenant_id - vault identifier
application - http
env - sandbox / live
message_type - Message Type Indicator (Hex Encoded)
execution_type - TCP execution type
label - CLIENT_TO_SERVER / SERVER_TO_CLIENT
Processing time for different phases of requests interacting with proxy, with breakdown on TCP message specifics. Duration - distributions of the amount of time each request takes.
iso_8583_proxy_request
(gauge)
tenant_id - vault identifier
application - http
env - sandbox / live
message_type - Message Type Indicator (Hex Encoded)
execution_type - TCP execution type
label - CLIENT_TO_SERVER / SERVER_TO_CLIENT
Representation of every HTTP or TCP request to measure. Rate - the number of requests, per minute, for the services that you are serving.
Custom Metrics
File Processing
file_processing_timestamps
state - file state
RECEIVED, PROCESSING, FAILED, PROCESSED, DELIVERED, UNDELIVERABLE (see the diagram below)
file_name - file name including the path prefix
tenant_id - vault identifier
File processing exposes a single metric called file_processing_timestamps with two dimensions: state and file_name and then a unix timestamp as the value. The timestamp is used to avoid any inaccuracy with timestamps as metrics are ingested
file_processing_bytes
state - file state
RECEIVED, PROCESSING, FAILED, PROCESSED, DELIVERED, UNDELIVERABLE (see the diagram below)
file_name - file name including the path prefix
tenant_id - vault identifier
Representation of the file size of the processed file
Files passing through VGS transition through several states that help provide insight into the progress of the file.
Observability Metrics data can be filtered along any of the dimensions in the table above.
Metrics can be used to help you understand the health of your integration. For example, if you notice the abnormal upstream responses, you can find the spike on the graph of non-200 status codes and all the details for failing requests. End-to-end visibility of latency and errors makes it easy to troubleshoot issues.
Example PromQL Queries
HTTP Proxy
View upstreams returning HTTP 5xx status codes
http_proxy_upstream_response{upstream_host=~".*upstreamhost.com", code=~"5.."}
This promql query will return any HTTP request to "*.upstreamhost.com" that has an HTTP status code starting with 5 (e.g., 500 - 599)
Last updated